- From: Stanton, John <StantonJ@ncr.disa.mil>
- Date: Sat, 1 Dec 2001 23:40:05 -0500
- To: www-webont-wg@w3.org
- Cc: "Swatski, Leonard" <swatskil@ncr.disa.mil>, "Miller, Kim A." <miller1k@ncr.disa.mil>, "Barnette, Jim" <barnettj@ncr.disa.mil>
Every use case expressed so far can provide leverage across the DOD in weapons; networks & Command & Control systems - especially when coupled with Agent Oriented Software Engineering. What the DoD always seems to need falls into two areas, in my judgment - security & formal methods that produce interoperable products. Formal methods is not so much meant to focus on formal methods of expression as in - context-free or context-dependent requirements expressed by narrative rules; or doing our own variant of Backus-Naur Form. What I mean is a connected; iterative standards development process that connects the standard under definition; the model; and a conformance testing system as they evolve TOGETHER. If all three of these elements evolve together in a clearly defined, intentional process, DOD can save money, and many other wondrous events can also occur. We purchase as much software as a Fortune 50 company. When we encounter a product that is 99% interoperable, the other 1% costs us millions of dollars to transport across platforms; or engineer expensive, weird work arounds that then require expensive life-cycle maintenance. When encountering this 1% non-interoperability, it can often be traced back to both the lack of formal methods of expression within the standard; but most often to the absence of an intentional overall standards development process, exploiting intentional software engineering, using iteration between the three major elements to produce quality; modeled; tested and evolved products. So... we suffer with these products having no way to test conformance; not understanding exactly what we have purchased, costing millions of dollars. I have prepared a 5 page paper titled, "A Generic QA Framework for Most any Standard Fora" and would happy to forward it after I get back from Oakland (Schema F2F), if there is interest. It might help me in explaining what the DoD really needs here. It is in MS/Word, and that's not being cool, I have been clearly told. Respectfully - John Stanton
Received on Saturday, 1 December 2001 23:40:20 UTC